Millivoltmeter - new meter scale

[retailer]

I'm working on a project that has an inbuilt millivolt meter, the ranges are switched in 10dB steps, it called for a meter with 0-1 and a 0-3 scale as well as a dB scale, looking though my meter collection I came across an old HP 1mA movement scrounged from a microwave power meter, it had the all scales that I needed, a stroke of luck I thought (without really looking closely at the meter) - no need to draw up a new scale. The project forged ahead until I took a close look at the meter, the HP scale was very non linear, no doubt to suit the HP circuit, a quick check with a variable resistor and 1.5 battery confirmed the worst, it needed to new scale drawn up.

Just in case other forum members are in need of the info needed to draw similar scales I'm posting, a screen shot of my excel worksheet calculated from +2dB to -20dB and the corresponding 1 and 3 (volt) scale - the number of divisions one puts in depends on the size of the meter, the HP meter is around 100mm square and I found that from -5dB on the scale gets cramped, it looked ok on the screen but once printed I could see it was no good and had to rethink how the mark it off so it was still legible and useable.

[Smithdoor]

I look at making a meter scale.
1) The manufacturers use screen printing great for hundreds.
2) print paper and glue on could give problems later if paper comes loss.
3) Etching on metal plate a lot of work .
4) Engravings costly
5) plating on new meter scale today some are plastic.
6) is what used but does not work every time is sharpie This does workbif building the meter you can adjust things to make right.

7) Is on I may try is go to place that does printing of photos and see if have a very stiff plastic they print on . This have not tried but looks good on paper.
8) Offset printer plates made of aluminum very costly
8. ) Offset printer plates made of aluminum very costly
9.) Iron on repaint meter scale and use heat transfer for new scale. This also works on pc boards and wood parts

Dave

PS Photo of sharpie used

[Radio Wrangler]

Many of HP's meters were individually scaled to suit the movement as well as to suit the measurement circuitry. Re-applying them for something else needs a bit of care. It might be worth feeding it a number of current values checked with a DMM to see how the pointer linearity is.

Some extreme meters had shaped pole pieces to give highly distorted scales. One that comes to mind was made to deflect proportionally to the logarithm of the current applied. This allowed a linear dB scale and consumed no power for an analogue computational system. Being logarithmic, with no drive, the springs take the pointer to the far left stop as an approximation for minus infinity dB !

They are superb movements but the white background paint can crinkle and fall off is subjected to prolonged humidity. I think Triplett made most of the movements. Most of the later rectangular style were taut-band suspended types.

David

[retailer]

I'm aware of meters with shaped pole pieces and did check this against my Fluke, and while there is some minor non linearity at the extreme ends, it's pretty much a regular moving coil meter. I started with a 1.5V battery and 10 turn potentiometer in series with the Fluke and marked off the divisions in 0.1mA increments on to the original scale with a pencil - scanned and imported it into Autocad so I could check it out, I used the Autocad polar array function to set lines at 0.1mA divisions and they are all either right on or within a few percent of the pencil marks I made.

It was the relationship between the 0-1, 0-3 and dB scales that threw me a bit and I had to go back to some text books and do some heavy (for me) reading before I worked it out, looking things up on the internet I think seems to be getting harder, you'd think the opposite would be true but time and time again the search engine link takes you to some text copied from another web site.

Not sure how the original meter scale was printed but I have a feeling it would have been silk screened - back when I had a few gap years from Uni I worked in the science labs at a local college and made various bits of gear during that time - always in lots of 8 for student use during practical lessons - once I had the metal work cut and drilled I'd wander over to the art centre and enlist the help of one of the teachers to make up silk screens using art work made with letraset, the results were always very good and I easily could do things as fine as a meter scale, it was a great place to work and I moonlighted making Quad 405 clones at the college during my spare time for a local music/PA hire company, a few days ago I came across a shoe box containing 405 PBC's and some 2N3773's I had left over from that time and it reminded me of those days .
I've made quite a few paper scales and had no issues with then coming loose or having glued let go, so that is what I'm going to do again, I usually stick it on with a bit of tape for a trial and once it checks out I'll glue it on permanently.

[retailer]

I decided to add 4,8 and16 ohm dummy loads that could be switched in or out and then thought a power into 8 ohms scale might come in handy so this will probably the final design to print.

[Radio Wrangler]

dBm is a power level relative to 1 milliwatt. So the positioning of the zero dBm marking to the voltage scale is set by what impedance the voltage is happening across. 223.6 mV RMS in a 50 Ohm system would be 0dBm. 774.6mV RMS in a 600 Ohm system would be 0dBm, and other voltages for other implied impedances.

dBm comes originally from the telephone industry. 1mW into the earpiece of a telephone, given their typical efficiency happens to be a good average signal level for audibility by the average user, and as it's a handy round number it got adopted as a standard reference level. 600m Ohms was the usual impedance in telephone systems, and human hearing has aroughly logarithmic perception of sound power, so a logarithmic scale of power is useful. Thus the dBm was born as a handy unit.

The RF world nobbled the dBm for its own uses, and it settled on two favourite impedances; 75 Ohms because it was the impedance of a typical half-wave dipole, and no-one's sure how 50 Ohms came about. It might just have been a round number close to where some early coax came out. So each of these impedances gives a different RMS voltage for its 0dBm reference. dBW might be more sensible in an SI sort of way, but dBm is strongly entrenched and because of human hearing, we have a natural feeling for what it means.

The white paint on which the meter markings is done on those HP meters is the problem. If I remember right, it is based on a product made from eggs! chosen for an easy and stable printing process, but susceptible to crack off over years if damp. Naval compass cards were made of rice paper because it was more temperature stable than normal wood fibre paper. There are all sorts of surprising materials kicking around in old instruments.

The meter scaling of 0-1 and 0-3.162 represent a 10dB change in power, so you'll see there is a slight difference in the position of 3 and 10. This allows voltage range switches to be labelled in 10dB steps, and the use of a 3.162 scale rather than 3.000 spread across the same angle as the 1.000 scale fixes that small difference. Everyone calls it a 0-1 and 0-3 scale but don't notice that extra bit above 3 on the meter face. So it's really a root-ten scale.

David

[emeritus]

I had a 1970's HP catalogue that explicitly mentioned the fact that the meter scales on at least one of its items of test gear were individually calibrated, and included photos of the scales of two different examples of the same model illustrating their differences. It might have been their vector voltmeter.

When I needed a meter with a scale covering 190V to 270V for use with an autotransformer, I drew two closely-spaced arcs in drawing ink on the piece of heavy card I was using for the scale using a compass with nib attachment, and marked the gradations with a fine Rotring pen, adjusting the voltage with a variac and an Avometer one lunchtime at work. Absolute precision was not necessary, 5V increments were more than adequate.

[DrStrangelove]

The lovely HP400EL has a linear dB scale.

I've long wanted one, but I've had to live with a 3400A instead.

The dBm scale on the 3400A has 0dBm aligned with 0.776V.